Abstracts in the order of presentations Estonian investigators´ questioning styles with child witnesses. Kristjan Kask* and Ray Bull Introduction It is found that open memory prompts tend to elicit longer and more accurate responses than do closed prompts. Reliance on open-ended questions during the early stages of interviews may not only produce greater amounts of uncontaminated information but also reduce acquiescence to misleading information introduced later in the interview. However, focused and closed questioning has been found to be dominant in forensic interviews of children in several countries. Aim The purpose of the present study is to examine the information provided by young children in Estonian police interviews of them. It is hypothesized that interviewers use more option-posing and direct questions than prompting children’s free recall and the proportions of usage of different questioning techniques will vary across time within interviews. Method Videotaped interviews with 17 children (mean age 8 years 10 months, range 5 to 13 years) were analyzed. The interviewers had received some formal training in interviewing but no special training for the questioning of child witnesses. Results A pattern of long interviewer questions and short child answers was often apparent. During the interview, with time the proportion of direct questions was found to decrease and the proportion of option-posing and suggestive questions to increase. Longer answers were provided in response to general or central invitations, whereas option-posing or closed questions produced less information. Conclusions When taking into consideration the frequency of the different interviewer utterance types, investigative interviews in Estonian sample share heavy reliance on directive and option-posing utterances found in the interviews conducted in other countries. Better police training in the adoption of best-practice guidelines in interviewing children is required. Therefore, more structured methods such as the NICHD protocol or cognitive interview are strongly recommended to structure the interviews, increasing both quality and quantity. Kristjan Kask University of Leicester School of Psychology Henry Wellcome Building Lancester Road Leicester LE1 9HN UNITED KINGDOM email: kk98@le.ac.uk Effects of collaboration with a non-witness on eyewitnesses’ recall correctness and metacognitive realism Farhan Sarwar*, Carl Martin Allwood, Åse Innes-Ker Institution for psychology, Lund University, Box 213, SE-221 00 Lund, Sweden Farhan.sarwar@psychology.lu.se Introduction Eye-witnesses to crimes often talk about the witnessed event with other persons. This issue is of great importance in forensic settings because of the effect that these retellings may have on the account of the event subsequently given by the witness in forensically important contexts. Aim To study the effects of interaction with a non-witness on eyewitnesses’ recall correctness and metacognitive realism. Method In this study the participants (n = 89) saw a short film of a staged kidnapping of a woman. The participants were randomly assigned to four conditions. In the lab-discussion condition the participants discussed the event five times in a two weeks period with a two-three days interval between each discussion and each time with a new confederate who asked non-prepared questions about the event. . In the rehearsal condition the participants only told the event to a confederate five times in a two weeks period. In the family discussion condition participants discussed the event with a new family member or friend five times in a two weeks period. In the control condition participants did nothing during the two weeks time. After the 2 weeks all participants first gave a free recall and then answered focused recognition questions and three days later gave confidence judgments of their previous memory recall. Results and conclusions Results for the free recall showed that the participants in the four conditions differed significantly in their accuracy, confidence and calibration but not on over-/underconfidence and in their ability to discriminate correct and incorrect memories by means of the level of their confidence judgments. No significant differences were found for the focused questions. Thus, it may be that the effects of witnesses’ communication with other persons about a witnessed event may have greater implications for features of their subsequent free recall compared with their answers to focused questions.